Fork me on GitHub
#boot
<
2016-07-11
>
escherize02:07:14

@myguidingstar: I found that occurrence very strange as well.

biscuitpants07:07:42

is it possible to call a task for function to happen just before quitting a running repl?

biscuitpants07:07:50

ie; just before an interrupt

martinklepsch08:07:14

@biscuitpants: check boot.core/cleanup

martinklepsch08:07:28

@escherize: did you try pinning clojure via boot.properties?

Petrus Theron09:07:47

Thanks, @sekao. It looks like the big difference is building cljs first, calling (target), then building clj and (target) again. Also the resource paths may affect it. Going to give it a try.

martinklepsch09:07:13

@petrus: regarding your error, do you have boot pinned via boot.properties?

martinklepsch09:07:29

@petrus: in general calling target twice to build a single artifact seems like it could be simplified

Petrus Theron09:07:11

@sekao, I鈥檓 trying to get the cljs+clj build working on Heroku, so I set env var BUILD_COMMAND=lein cljsbuild once && lein uberjar, which throws java.lang.Exception: Unknown build identifier: &&

Petrus Theron09:07:01

@martinklepsch, I have boot pinned via BOOT_VERSION=2.6.0 in boot.properties.

Petrus Theron09:07:27

Whoops, that wasn鈥檛 meant for me 馃檪

sekao10:07:59

@petrus: i don鈥檛 know much about heroku, sorry. you might need to somehow quote that command.

Petrus Theron10:07:31

Thanks, got it working with a lein alias: :aliases {"go" ["do" ["cljsbuild" "once" "production"] ["uberjar"]]}

Petrus Theron10:07:44

still fighting

anmonteiro15:07:33

I鈥檝e now run into this issue a couple of times

anmonteiro15:07:52

only solved with :ensure-clean false, even though my repo is clean

anmonteiro15:07:02

I don鈥檛 know what else to be looking for

alandipert15:07:35

@anmonteiro: git status showed nothing to commit, working directory clean and you still got the error about clean?

anmonteiro15:07:11

On branch master
Your branch is up-to-date with 'origin/master'.
nothing to commit, working directory clean

anmonteiro15:07:29

@alandipert: and I also suspect it鈥檚 what the reporter of that issue had too

alandipert15:07:28

maybe something about the repo is confusing jgit?

anmonteiro15:07:55

@alandipert: yeah, I also had a look at that

alandipert15:07:03

could play around with it at the repl and see what it's doing

alandipert15:07:12

ie (in-ns 'boot.jgit)

anmonteiro15:07:15

I suppose I could do that

anmonteiro15:07:50

@alandipert: that won鈥檛 work

anmonteiro15:07:59

also can鈥檛 (require 鈥榖oot.jgit)

anmonteiro16:07:31

doesn鈥檛 seem to be on the classpath, somehow

anmonteiro16:07:47

I haven鈥檛 used those ever, sorry

anmonteiro16:07:48

炉\(銉)/炉

martinklepsch16:07:23

I think if you depend on [boot/worker "2.6.0"] in a repl it would work

martinklepsch16:07:47

not 100% sure if that's the right module but should be

anmonteiro16:07:17

@martinklepsch: that would mean my repo would not be clean anymore

anmonteiro16:07:27

I don鈥檛 really want to commit that to my history too

alandipert16:07:34

sorry, boot repl -P worker will get you into the worker pod where jgit stuff lives

anmonteiro16:07:43

cool, thanks! let me try that

martinklepsch16:07:11

@anmonteiro: you can always set-env! in a repl

anmonteiro16:07:20

found the culprit

(status)
{:added #{}, :changed #{}, :missing #{}, :modified #{}, :removed #{}, :untracked #{".DS_Store"}}

anmonteiro16:07:47

it doesn鈥檛 show in magit or git status however

alandipert16:07:58

ah, maybe it's ignored in oyur global git prefs

anmonteiro16:07:13

it probably is

alandipert16:07:34

my guess is adding it to the project gitignore will fix

anmonteiro16:07:35

@alandipert: how could I test this change? which value should I use if I want to push to the local repo in the push task?

alandipert17:07:57

@anmonteiro: it depends on whether you want it to let you push even if git is dirty

alandipert17:07:42

@anmonteiro: if you want that, then for it to not see my theory is you need to add .DS_Store to your project's .gitignore

anmonteiro17:07:38

@alandipert: right, I did know that. What I鈥檓 asking is how could I test after making the change of putting that in gitignore

anmonteiro17:07:48

I dont鈥 really want to push it to clojars again

alandipert17:07:02

try (status) at the repl again? should be empty

anmonteiro17:07:09

yeah, makes sense, sorry for the noise 馃檪

alandipert17:07:29

np ! hope it works

alandipert17:07:55

connecting to pods via repl is boot wizard level 3.4 stuff btw

alandipert17:07:01

congrats on leveling up

anmonteiro17:07:12

ahah, thanks 馃檪